Multi-View Relational Distillation for Spatial Reasoning with Vision-Language Models

• URL: http://arxiv.org/abs/2608.10864v1
• Date: Tue, 11 Aug 2026 12:34:41 GMT
• Title: Multi-View Relational Distillation for Spatial Reasoning with Vision-Language Models
• Abstract summary: Vision-language models (VLMs) have achieved strong image and video understanding, yet their visual-spatial representations remain geometrically fragile.<n>We propose multi-view relational distillation (MVRD), which distills patch-wise cosine similarities across views instead of the teacher features themselves.
• Score: 20.202625597458177
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Vision-language models (VLMs) have achieved strong image and video understanding, yet their visual-spatial representations remain geometrically fragile, leading to failures in spatial reasoning needed for embodied AI, robotics, and autonomous driving. Prior approaches to geometry grounding either fine-tune VLMs on spatial question answering, which can perpetuate spurious visual representations, or fuse features from large geometry-grounded vision models, which substantially increases model size at inference. Knowledge distillation from geometry-grounded vision models offers an alternative, but directly matching multi-view teacher features can disrupt the pretrained alignment between visual and textual representations, degrading object- and language-semantic capabilities. We propose multi-view relational distillation (MVRD), which distills patch-wise cosine similarities across views instead of the teacher features themselves. These relations encode geometric correspondences adequate for spatial understanding, while leaving the student representation underdetermined, allowing it to remain close to its pretrained vision- language space. Across representative VLMs, MVRD improves visual-spatial reasoning, outperforming supervised fine-tuning and feature distillation while approaching feature fusion methods with considerably fewer added parameters and lower latency. We show that MVRD makes visual representations more geometric while retaining language alignment, and generalizes to 3D scene understanding tasks such as object grounding, dense captioning, and question answering.

Related papers

• Which Pretraining Paradigm Better Serves Spatial Intelligence? An Empirical Comparison of Vision-Language and Video Generation Models [31.851717131965824]
  We present the first systematic frozen-feature probing study of Vision-Language Models (VLMs) and Video Generation Models (VGMs)<n>Using the lightweight probe, our framework enables a controlled comparison of what information is already encoded in frozen representations from two model families.<n>Results reveal a clear complementarity: VLMs are stronger at semantic tagging and instance grouping, while VGMs provide more accessible signals for dense geometry and camera motion.
  arXiv  Detail & Related papers  (2026-05-27T08:20:04Z)
• SpatialImaginer: Towards Adaptive Visual Imagination for Spatial Reasoning [67.67774742200626]
  spatial intelligence, which refers to the ability to reason about geometric and physical structure from visual observations, is a core challenge for large language models.<n>We propose a unified multimodal generation framework that integrates textual reasoning with visual imagination.<n>Our framework adopts a divide-and-conquer strategy, using text chain-of-thought for high-level semantic planning and the visual imagination for geometry-sensitive state transformation and consistency preservation.
  arXiv  Detail & Related papers  (2026-04-19T11:21:59Z)
• Learning 3D Representations for Spatial Intelligence from Unposed Multi-View Images [81.94999489820974]
  UniSplat is a feed-forward framework for learning 3D representations from unposed multi-view images.<n>We introduce a dual-masking strategy that strengthens geometry induction in the encoder.<n>Second, we develop a coarse-to-fine Gaussian splatting strategy that reduces appearance-semantics inconsistencies.<n>Third, we introduce a pose-conditioned recalibration mechanism that interrelates predicted 3D point and semantic maps into the image plane.
  arXiv  Detail & Related papers  (2026-04-12T10:36:18Z)
• Seeing through Imagination: Learning Scene Geometry via Implicit Spatial World Modeling [68.14113731953971]
  This paper introduces MILO, an Implicit spatIaL wOrld modeling paradigm that simulates human-like imagination.<n>We show that our approach significantly enhances spatial reasoning capabilities across multiple baselines and benchmarks.
  arXiv  Detail & Related papers  (2025-12-01T16:01:41Z)
• Chain-of-Visual-Thought: Teaching VLMs to See and Think Better with Continuous Visual Tokens [54.18057944158818]
  Chain-of-Visual-Thought (COVT) is a framework that enables Vision-Language Models (VLMs) to reason through continuous visual tokens.<n>Within a small budget of roughly 20 tokens, COVT distills knowledge from lightweight vision experts.<n>During training, the VLM with COVT autoregressively predicts visual tokens to reconstruct dense supervision signals.
  arXiv  Detail & Related papers  (2025-11-24T18:55:19Z)
• Visual Jigsaw Post-Training Improves MLLMs [58.29961336087896]
  We introduce Visual Jigsaw, a generic self-supervised post-training framework designed to strengthen visual understanding in large language models (MLLMs)<n>Visual Jigsaw is formulated as a general ordering task: visual inputs are partitioned, shuffled, and the model must reconstruct the visual information by producing the correct permutation in natural language.<n>Extensive experiments demonstrate substantial improvements in fine-grained perception, temporal reasoning, and 3D spatial understanding.
  arXiv  Detail & Related papers  (2025-09-29T17:59:57Z)
• Evo-0: Vision-Language-Action Model with Implicit Spatial Understanding [11.222744122842023]
  We introduce a plug-and-play module that implicitly incorporates 3D geometry features into Vision-Language-Action (VLA) models.<n>Our method significantly improves the performance of state-of-the-art VLA models across diverse scenarios.
  arXiv  Detail & Related papers  (2025-07-01T04:05:47Z)
• Mind the Gap: Benchmarking Spatial Reasoning in Vision-Language Models [14.442394137843923]
  We present a detailed analysis that first delineates the core elements of spatial reasoning.<n>We then assesses the performance of these models in both synthetic and real-world images.
  arXiv  Detail & Related papers  (2025-03-25T14:34:06Z)
• RelationVLM: Making Large Vision-Language Models Understand Visual Relations [66.70252936043688]
  We present RelationVLM, a large vision-language model capable of comprehending various levels and types of relations whether across multiple images or within a video. Specifically, we devise a multi-stage relation-aware training scheme and a series of corresponding data configuration strategies to bestow RelationVLM with the capabilities of understanding semantic relations.
  arXiv  Detail & Related papers  (2024-03-19T15:01:19Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.